BODROSSY, Levente (Deák Ferenc u.1, Töltéstava, H-9068, HU)
KOSTIC, Tanja (Heinestrasse 22/20, Vienna, A-1020, AT)
WEILHARTER, Alexandra (Wehlistrasse 51/2/43, Vienna, A-1200, AT)
SESSITSCH, Angela (Mittelstrasse 48, Vienna, A-1140, AT)
BODROSSY, Levente (Deák Ferenc u.1, Töltéstava, H-9068, HU)
KOSTIC, Tanja (Heinestrasse 22/20, Vienna, A-1020, AT)
WEILHARTER, Alexandra (Wehlistrasse 51/2/43, Vienna, A-1200, AT)
| - -
Claims :
1. A set of oligonucleotides specific for the gyrB gene of a microorganism, characterized in that at least one oligonucleotide comprises a sequence of SEQ ID NOs 1 to 71 or a fragment of a sequence of SEQ ID NOs 1 to 71 of at least 14 nucleotides, with optionally 1 or 2 point mutations, and complementary reverse sequences thereto.
2. The set of claim 1, characterized in that at least one oligonucleotide comprises a full sequence of SEQ ID NOs 1 to 71.
3. The set of claim 1 or 2, characterized in that each oligonucleotides is specific for the gyrB gene of one microorganism and the set comprises oligonucleotides specific for at least two microorganisms of different genera.
4. The set according to any one of claims 1 to 3, characterized in that the oligonucleotides comprise sequences selected from at least two, more preferably at least 10, most preferably at least 30 different sequences of SEQ ID NOs 1 to 71.
5. The set according to any one of claims 1 to 4, characterized in that the set of oligonucleotides consists of the complementary reversed oligonucleotides.
6. The set according to any one of claims 1 to 5, characterized in that the oligonucleotides are labeled.
7. The set according to any one of claims 1 to 6, characterized in that the sequences are of SEQ ID NOs 1 to 69.
8. A solid support wherein the oligonucleotides of the set of any one of claims 1 to 7 or oligonucleotides which are complementary reversed to the oligonucleotides of the set of any one of claims 1 to 7 are immobilized on the solid support.
9. The solid support according to claim 8, characterized in that different oligonucleotides are each immobilized in different spots on the solid support. - / B ~
10. The solid support according to claim 8 or 9, characterized in that it forms a microarray.
11. The solid support according to any one of claims 8 to 10, characterized in that the solid support comprises an aldehyde surface.
12. A method for identifying a microorganism, comprising i) providing DNA of the gyrB gene of the microorganism, ii) hybridizing the gyrB DNA with the set according to any one of claims 1 to 7 or the solid support according to any one of claims 8 to 11, wherein the oligonucleotides form immobilized probes, under stringent conditions, iii) optionally labelling hybridized oligonucleotides, iv) visualising hybridized oligonucleotides.
13. A method for identifying a microorganism, comprising a) providing the set of oligonucleotides according to any one of claims 1 to 7 specific for the gyrB gene of the microor- gansim, b) providing a support with immobilized probes, the probes being capable of hybridizing to the oligonucleotides under stringent conditions, c) providing a sample of microorganism DNA or RNA, which is preferably amplified by PCR, most preferred PCR specific for the gyrB gene, d) hybridizing the oligonucleotides with the sample under stringent conditions, e) labelling hybridized oligonucleotides, f) hybridizing the oligonucleotides against the probes on the support under stringent conditions, g) visualising the labelled oligonucleotides bound to the probes .
14. The method according to claim 13, characterized in that the probes comprise the reverse complement sequences of the oligonucleotides, and preferably a linker for binding to the support.
15. The method according to any one of claims 12 to 14, charac- terized in that the support and the immobilized probes form a microarray.
16. The method according to any one of claims 12 to 15, characterized in that the labelling is performed by addition of at least one labelled ddNTP, preferably ddCTP.
17. The method according to any one of claims 12 to 16, characterized in that during a hybridization competitive oligonucleotides are present, which differ from the oligonucleotides of the set by at least one mismatch, preferably 1 to 6 mismatches, most preferred 1 to 3 mismatches.
18. The method according to claim 17, characterized in that the competitive oligonucleotides cannot bind a label, preferably have a 3' phosphate modification incompatible with a nucleotide addition reaction.
19. The method according to claim 17 or 18, characterized in that the competitive oligonucleotides comprise at least one, preferably at least 3, most preferred at least 6 different sequences selected from SEQ ID NOs 74 to 84.
20. The method according to any one of claims 12 to 19, characterized in that the microorganism is a pathogen.
21. The method according to claim 20, characterized in that the the pathogen is selected from E. coli, Shigella spp, Salmonella spp, A. hydrophila, V. cholerae, M. avium, M. tuberculosis, H. pylori, P. mirabilis, Y. enterocolitica and C. jejuni.
22. The method according to any one of claims 12 to 21, characterized in that steps i) to iv) or a) to g) of a first analysis are repeated in at least a second analysis with a set of oligonucleotides excluding oligonucleotides for microorganisms used in a previous analysis.
23. The method according to any one of claims 12 to 22, characterized in that at least 10, preferably at least 30, most preferred at least 70 different probes are immobilized on the support.
24. The method according to any one of claims 12 to 23, characterized in that the probes on the support comprise sequence (s) that is/are selected from at least one, more preferably at least 10, most preferably at least 30 different sequences that are complementary reversed to SEQ ID NOs 1 to 71. |
Set of oligonucleotides
The present invention relates to the microbiological field of nucleotide based microorganism detection and identification.
Microarrays are genomic tools originally developed to monitor gene expression, also applied for the detection of specific mutations in DNA sequences and lately employed in the parallel detection and identification of microorganisms in environmental or clinical samples. Although the use of microbial diagnostic microarrays (MDMs) has increased lately (Loy et al., 2002; Bo- drossy et al., 2003; Lee & Chao, 2005; Hashsham et al., 2004; Tiquia et al., 2004; DeSantis et al., 2005; Schadt et al., 2005; Bae et al., 2005; Wang et al., 2005; Lehner et al . , 2005; Lein- berger et al., 2005; Neufeld et al., 2006), application on environmental scale still presents great challenges in terms of specificity, sensitivity and quantification (Zhou, 2003) . MDM sensitivity can be defined in different ways. The amount of nucleic acid needed for successful detection; the ratio of microbial nucleic acid to background, host nucleic acid; and the ratio of target to non-target microbial nucleic acid. The third type of . sensitivity threshold, directly related to the relative abundance of the target microbe (s) within the microbial community, is the one most frequently limiting the applicability of MDMs. The current reported sensitivity threshold of MDMs lies in the range of 1 to 5% relative abundance (Bodrossy et al . , 2003; Denef et al., 2003; Tiquia et al . , 2004). MDMs for microbial community analysis need to be capable of detecting microbes from a broad phylogenetic range. In turn, their application can tolerate a relatively low sensitivity in terms of relative abundance. On the other hand, high specificity and sensitivity are primary requirements for MDMs applied for pathogen detection. Such MDMs, used in clinical, veterinary, food and public health microbiology, typically rely on species or genus specific PCR amplifications to increase sensitivity and specificity of detection (Sergeev et al., 2004; Lee & Chao, 2005; Maynard et al . , 2005) . As a drawback, they are then limited to a narrow range of pathogens. To cover a range of different pathogens, this approach requires multiple PCR or multiplex PCR (PCR with multiple primer pairs at once) reactions to be performed. Epidemiological and public health surveys as well as the screening of veterin-
ary, food or water samples for pathogens requires a different approach, combining broad coverage with high specificity and increased sensitivity.
The CN 1396270 describes a DNA rαicroarray for detecting the frequently encountered pathogenic bacteria in water, such as Escherichia 's bacteria, salmonella, staphylococcus, etc., which uses the lβs rRNA gene, two primers designed in the preservation region of 16s rRNA gene, and a probe for the variable region of lβs rRNA gene.
The DE 199 45 916 describes several oligonucleotide sequences for bacterial 23S/5S rRNA probes, which can be used in assays to distinguish bacteria of different categories.
The WO 2004/046379 A describes the identification of microorganisms through variable regions of topoisomerase genes.
A goal of the present invention is to provide oligonucleotide probes capable of the sensitive and specific detection of a broad range of microorganisms from samples harbouring complex microbial communities.
The present invention provides a set of oligonucleotides specific for the gyrB gene of a microorganism, characterized in that at least one oligonucleotide comprises a sequence of SEQ ID NOs 1 to 71, preferably of SEQ ID NOs 1 to 69, or a fragment of a sequence of SEQ ID NOs 1 to 71, preferably of SEQ ID NOs 1 to 69, of at least 14 nucleotides (preferably up to 30 nucleotides) , with optionally 1 or 2 point mutations, and complementary reverse sequences thereto. In this context preferred are the full nucleotides of the given sequences, but also fragments with lengths of at least 8, 10 , 12, 14, 16, 20, 24 or 28 nt lengths are possible. The sequences may also comprise further nt sequences or not. The mutations allow the detection of other closely related microorgansims leading to mismatches in hybridization experiments with sample genetic material, or amplified material of one or more microorganism (s) . The suitability of such derivatives can be steered by the skilled man in the art when using the inventive set by varying standard assay conditions modifying stringency. In the context of point mutation a nucleotide exchange but also a deletion or an insertion is possible. Particularly preferred the mutations are at the 3' end of the oligonucleotide sequences. However preferred are the sequences given in SEQ ID NO 1 to 71.
- -
The present invention also provides a set of oligonucleotides specific for the gyrB gene of at least two microorganisms of different genera, which can be used for the method described herein.
Preferably the oligonucleotides comprise sequences selected from at least two, more preferably at least 10, most preferably at least 30 different sequences of SEQ ID NOs 1 to 71. The sequences of this set are specific for the gyrB gene of different species or genera of selected microorganisms and can be used for their detection or identification. Depending on different methods, e.g. hybridization methods, PCR based or microarray methods using different stringency conditions each oligonucleotide can detect only one microorganism or closely related microorganisms (e.g. via lower stringency).
Preferably the set comprises oligonucleotides, wherein each oligonucleotides is specific for the gyrB gene of one microorganism and the set comprises oligonucleotides specific for at least 2, 3, 4, 5, β, 7, 8, 9, 10, 12, 16, 24 or 32 microorganisms of different genera, thus allowing a broad scope of detectable microorganisms .
Preferably the set comprises oligonucleotides, wherein different oligonucleotides comprise sequences selected from at least 2, 3, 4, 5, 6, 7, 8, 9, 10, 12, 16, 24, 30 or 32 different sequences of SEQ ID NOs 1 to 71.
In another aspect a set of oligonucleotides is provided, comprises of oligonucleotides, which are complementary reversed to the oligonucleotides of a sequence of SEQ ID NOs 1 to 71. Such oligonucleotides, which sequence is complementary reversed to a template oligonucleotides can thus hybridize these template oligonucleotides. Complementary reversed oligonucleotides are also specific for the anti-sense strand of a genetic (or amplified) DNA strand. These sets may also comprise further oligonucleotides, especially microorgansim specific oligonucleotides, e.g. directly selected from SEQ ID NOs 1 to 71, thus forming mixtures in the specificity for the gyrB strand in sense and anti-sense direction.
Preferably oligonucleotides of the set are labeled.
In another aspect the present invention provides a solid support, wherein oligonucleotides of the set or oligonucleotides which are complementary reversed to the oligonucleotides of the
set (and can thus hybridise the oligonucleotides of the set) are immobilized on the solid support.
In another aspect the present invention provides a solid support comprising at least two probes specific for the gyrB gene of at least two microorganisms of different genera immobilized on the solid support.
Also preferred, the probes on. the support comprise sequences that are selected from at least two, more preferably at least 10, most preferably at least 30 different sequences that are complementary reversed to SEQ ID NOs 1 to 71 (and therefore can hybridise these sequences) .
Advantageously, said oligonucleotide molecules are immobilized in different spots on the surface. By providing spots, specifically defined regions comprising the immobilized oligonucleotide molecules are produced. In case one spot is specific for certain microorganism it is possible to exactly define the microorganism present in the sample analysed. It is further possible to provide a reference oligonucleotide molecule in the spots which is not only a positive control but can also be used as a reference for quantification.
Preferably the solid support forms a microarray. DNA mi- croarrays are a powerful tool for the parallel, high-throughput detection and quantification of many genes. Originally developed for whole genome gene expression analyses, DNA microarrays have a very strong application potential in many areas of microbiology. Upon availability of corresponding probe sets, they enable the detection of up to several thousand microbial strains, species or groups of species (depending on the design of the probe) in a single assay. In clinical, veterinary and plant microbiology, food and water quality control, this means that a single test can be developed to detect all pathogenic/beneficial/contaminating bacteria which might be present in the investigated sample. The potential for environmental microbiology is even stronger. By applying nested sets of oligonucleotide probes which target genes reflecting the phylogeny of the carrying organism, it becomes possible to roughly assess the whole proka- ryotic diversity of an environment.
A microarray comprises a large number of immobilized oligonucleotide molecules provided in high density on the solid support. A microarray is a highly efficient tool in order to detect
hundreds to millions of target molecules in one single detection step. Such microarrays are often provided as slides or plates in particular microtiter plates. In the state of the art a microar- ray is both defined either as a miniaturized arrangement of binding sites (i.e. a material, the support) or as a support comprising miniaturized binding sites (i.e. the array) . Both definitions can be applied for the embodiment of the present invention. For the first of these definitions the preferred embodiment of the present invention is a miniaturized arrangement of the oligonucleotides of the present invention in a microarray. The oligonucleotide molecules are preferably immobilized onto the microarray with the help of a printing device which ensures immobilization in high density on the solid support. This microarray is particularly useful when analysing a large number of samples .
Especially preferred the solid support comprises an aldehyde surface. An aldehyde surface can be used for the immobilisation of oligonucleotide probes, especially with a NH 2 -group at one end, preferably included by the linker. The linker is preferably a Ci to Cio, preferably a C 2 to C 6 linker bound to a short, e.g. 4, 5, 6 nt long, T-stretch.
In another aspect the present invention provides a method for identifying a microorganism, comprising i) providing DNA of the gyrB gene of the microorganism, ii) hybridizing the gyrB DNA with the set of oligonucleotides of the present or the solid support, wherein the oligonucleotides form immobilized probes, under stringent conditions, iii) optionally labelling hybridized oligonucleotides, iv) visualising hybridized oligonucleotides.
Through the specific hybridisation with the set of the present invention a multitude of microorgansims can be detected and identified via the gyrB gene. The specificity of MDMs is predominantly defined by the degree of conservation of the marker gene and the length of the oligonucleotide probe (Loy & Bo- drossy, 2006) . The gyrB gene, encoding the subunit B of the bacterial gyrase meets all the requirements for a phylogenetic- ally useful protein-coding gene: it can be found in most bacterial species and it does not appear to be frequently horizontally transmitted. Furthermore, the rate of gyrB evolu-
tion is not only faster than that of the ribosomal genes, but also appears fast relative to other protein coding housekeeping genes (http://seasquirt.mbio.co.jp) (Yamamoto & Harayama, 1995). Short oligonucleotide probes enable the discrimination of single nucleotide differences. In combination with a high resolution phylogenetic marker, an optimal performance can be achieved in terms of MDM specificity. A method according to the present invention is capable of the sensitive and specific detection of a broad range of microorganisms from samples harbouring complex microbial communities.
Preferred stringent conditions are 60 0 C annealing temperature for labelling, and 55 0 C for hybridization with the probes of the solid support, with βxSSC, lxDenhardt's reagent (Sigma, St. Louis, MO, USA), 0,1% SDS, especially in cases where the reactive oligonucleotides are designed for a melting temperature of 60 0 C (e.g. +/- 2 0 C, however a significant proportion of oligos with lower Tm can still work) , for both the hybridization of the reactive oligonucleotides (RC oligos) to the sample (e.g. the PCR product) and/or the probes on the support. However oligonucleotides with different overall melting temperatures can also be designed, e.g. 55°C by selecting gyrB stretches with different GC content, thus lowering the preferred stringent temperature to the melting temperature. In especially preferred cases the stringency is increased by increasing the designed melting temperature by 5 0 C above the average melting temperature of the oligonucleotides .
In a further embodiment a method for identifying a microorganism is provided, comprising a) providing the inventive set of oligonucleotides specific for the gyrB gene of the microorgansim, b) providing a support with immobilized probes, the probes being capable of hybridizing to the oligonucleotides under stringent conditions, c) providing a sample of microorganism DNA or RNA, which is preferably amplified by PCR, most preferred PCR specific for the gyrB gene, d) hybridizing the oligonucleotides with the sample under stringent conditions, e) labelling hybridized oligonucleotides, f) hybridizing the oligonucleotides against the probes on the
support under stringent conditions, g) visualising the labelled oligonucleotides bound to the probes .
In a preferred embodiment the probes comprise the reverse complement sequences of the oligonucleotides (to allow the hybridization to the oligonucleotides), and preferably a linker for binding to the support.
Preferably the support, preferably solid support, and the immobilized probes form a microarray.
In a preferred embodiment the probes are oligonucleotide probes, in particular DNA or RNA with lengths between 10 to 30 nts.
In another preferred embodiment the labelling is performed by addition of at least one labelled ddNTP. The addition of ddNTPs prevents further labelling and signal alteration. Preferably the label is a fluorescence label or a radioactive label, e.g. 32 P. Preferably the labeled nucleotides are ddCTP.
Preferably during a hybridization competitive oligonucleotides are present, which differ from the oligonucleotides of the set by at least one mismatch, preferably 1 to 6 mismatches, most preferred 1 to 3 mismatches. These competitive oligonucleotides (CO oligos) can bind to the sample with less specificity than a respective RC oligo and help to reduce or prevent false positive signals.
In a special embodiment the competitive oligonucleotides cannot bind a label, preferably have a 3' phosphate modification incompatible with a nucleotide addition reaction (e.g. sequence specific end labeling) - thus the competitive oligonucleotides cannot bind a label.
In preferred embodiments the competitive oligonucleotides comprise at least one, preferably at least 3, most preferred at least 6 different sequences selected from SEQ ID NOs 74 to 84, given in table 3 below.
Preferably, the microorganism is a pathogen, in particular selected from E. coli, Shigella spp, Salmonella spp, A. hydro- phila, V. cholerae, M. avium, M. tuberculosis, H. pylori, P. mirabilis, Y. enterocolitica and C. jejuni.
Especially preferred the steps (e.g. steps i) to iv) or a) to g) of the methods above) of a first analysis are repeated in at least a second analysis with a set of oligonucleotides ex-
eluding oligonucleotides for microorganisms used in a previous analysis. Especially with environmental samples this improves the detection and identification of less abundant microbes within a complex microbial community.
Preferably, at least 4, 9, 10, 12, 16, 18, 20, 24 or 26, preferably at least 30, 36 or 40, most preferred at least 50, 60 or 70 different probes are immobilized on the support.
Also preferred is that the oligonucleotides comprise sequence (s) selected from at least one, more preferably at least 4, 9, 10, 12, 16, 18, 20, 24 or 26, preferably at least 30, 36 or 40, most preferred at least 50, 60 or 70 different sequences of SEQ ID NOs 1 to 71, given in table 1 below.
In another embodiment the probes on the support comprise sequence (s) that is/are selected from at least 4, 9, 10, 12, 16, 18, 20, 24 or 26, preferably at least 30, 36 or 40, most preferred at least 50, 60 or 70 different sequences that are complementary reversed to SEQ ID NOs 1 to 71, given in table 1 below.
The present invention is further illustrated by the following figures and examples.
Figures:
Fig. 1: Probe set validation.
Predicted (weighted mismatches values as calculated with CaI- cOligo 2.03) and experimentally established probe specificity.
Fig. 2: Sensitivity of detection.
Microarray images showing hybridisation results from a soil DNA sample (A) and from the same soil DNA, spiked with 0.1% Vibrio cholerae (B) . Each array contains the same probes in triplicates. Images were scanned at 100% laser power, 750 V PMT and are displayed in rainbow colour mode. Setting for brightness was 78% in both cases; for contrast 78 and 81%, respectively. Normalised signal values for Vch_1776 and Vch_1839 were 4.6 and 7.2, respectively, corresponding to around 3% of the maximal signal obtained for them with pure cultures (during validation). 1: Internal control Msi_294. 2: Vch_1776 3: Vch_1839.
Fig. 3.: Application of competitive oligonucleotides. Microarray images showing hybridisation results of the En-
terobacter cloacae labelled without (A) and with (B) competitive oligonucleotides. Each array contains the same probes in triplicates. Images were scanned at 100% laser power, 750 V PMT and are displayed in rainbow colour mode. 1: Internal control Msi_294. 2: Eco_1402 and 3 - Eco_1404 false positive signals that could be successfully silenced with the addition of the competitive oligonucleotide C0_2. 3: Yer_1740 false positive signal that could be successfully silenced with the addition of the competitive oligonucleotides C0_7
Fig. 4: Sequence specific end-labelling of oligonucleotides (SSELO) principle.
E x a m p l e s :
Detection and identification methods for pathogens need in many cases detect less abundant microbes within a complex microbial community. In addition, a high specificity is required, providing robust, reliable identification at least at the species level. A microbial diagnostic microarray approach, combining a unique labelling method, gyrB as the marker gene, and the application of competitive oligoprobes was developed, enabling the sensitive and specific detection of a broad range of pathogenic bacteria from samples harbouring complex microbial communities. The approach was first tested with a set of 35 oligoprobes targeting E. coli, Shigella spp, Salmonella spp, A. hydrophila, V. cholerae, M. avium, M. tuberculosis, H. pylori, P. mirabilis, Y. enterocolitica and C. jejuni. The introduction of competitive oligonucleotides in the labelling reaction successfully suppressed cross-hybridisations by closely related sequences. Environmental applicability was tested with environmental and veterinary samples. A detection sensitivity in the range of 0.1% has been demonstrated.
Example 1: Materials and methods
Probe design and microarray fabrication
The gyrB sequence database was established by downloading sequences from the NCBI database (http://www.ncbi.nlm.nih.gov) and by the sequencing of the strains used for microarray validation. Alignment and neighbour joining phylogenetic tree of the gyrB sequences were constructed using the ARB software package
(Ludwig et al., 2004), which was subsequently also used for the probe design. The most important point considered during probe design was the placement of the diagnostic mismatch (es) as close to the 3' end as possible. This feature is very important for the sequence-specific end labelling of the oligonucleotides. The probes are designed in a way, that:
• They are 100% matching the target organism
• Mismatches of non-targeted, but related sequences (those, which have one, two or three mismatches against the probe) are preferably located towards the 3' end of the probe
In the next sequence example the oligonucleotide is underlined (it is presented in a 5' => 3' direction) . Below that are listed the respective sequences from the target (Salmonella) and some non-target, but related bacteria. Each non-target organism has a mismatch at the 3' position (highlighted in italic) . Some probes have mismatches further away from the 3' end. More like those highlighted in bold - even though the 3' mismatches are already enough to guarantee specificity.
RC oligo: Sal_1451 name fullname mis N_mis wmis pos ecoli rev ' AGGCACCCCOGGCCGϋC'
*Syhim gy00019:S.typhimurium 0 0 0.0 1451 1451 0 GCGUGCCGC-=================-ACOGGCGAϋ
*Etrbc Enterobacter_cloacae 2 0 1.7 1451 1451 0 GCGUGCCGC-=======G========ti-ACCGGCGAU
*Klbsl Klebsiella_pneumoniae 2 0 2.2 1451 1451 0 GUGϋGCCGC-=======G========G-ACUGGCGAA
*EslYyy23 Escherichia_coli 3 0 2.8 1451 1451 0 GϋGUACCGC-=======G=====G==u-ACCGGCGAG
*Esrchl34 Escherichia_coli 4 0 3.9 1451 1451 0 GϋGUACCGC-====O==G=====G==u-ACCGGCGAG
*Shgll.l Shigella_sonnei 4 0 3.4 1451 1451 0 GϋGUACCGC-=======Gu====G==u-ACCGGCGAG
*Erbct_5 Enterobacter_cloacae 2 0 1.7 1451 1451 0 GCGσGCCGC-=======G========u-ACUGGCGAA
*SnsMar Serratiajnarcescens 4 0 3.9 1451 1451 0 GCGAGCCGC-====g==G===A====G-GUGGGCGAA
Furthermore, all probes were designed with a 3' terminal cytosine residue that was added via terminal labelling of the oligonucleotides (upon the presence of a corresponding template) . Other factors considered during probe design were similar melting temperature Tm (targeted 60 0 C), and length between 17 and 28 nucleotides. Outputs of the Probe Match function were imported into CalcOligo 2.03 (Stralis-Pavese et al . , 2004) . CalcOligo was used to create an Excel table indicating predicted melting temperatures (based on the nearest neighbour model and SantaLucia parameters (SantaLucia, Jr. et al., 1996)),
length and GC content of the probes and the number of weighted mismatches between each probe-target pair. Nearest neighbour Tm values were calculated with concentration settings of 250 nmol for oligonucleotide and 50 mmol for Na+. Factors for weighing mismatches in CalcOligo were as follows. Positions: 5' 1st 0.3; 5' 2nd 0.6; 5' 3rd 0.8; 3' 1st 4.0; 3' 2nd 2.0; 3' 3rd 1.2; all other positions 1.0. Basepairs dArC 1.2; dTrC 1.2; dGrϋ 0.7; dTrG 0.4; all other mismatched basepairs 1.0 (where λ M" refers to the probe on the array and "r" to the target sequence) . Probe-target pairs with weighted mismatch values of up to 0.5 were expected to yield positive hybridisation under the conditions applied. Complete list of the oligonucleotides used in this study can be found in Table 1. Control oligonucleotide
(Msi_294) targeting pmoA gene of the Methylosinus trichosporium OB3b was taken from the diagnostic microarray for methanotrophs
(Bodrossy et al., 2003).
Table 1: Oligonucleotide probe set. Sequences listed are those of the RC oligo probes that were used in the labelling reaction. All sequences are listed without the 3' terminal C residue. Indicated Tm and G+C % values were calculated using CalcOligo 2.03. SEQ ID Nr are given for PCR reaction oligonucleotides ("RC oligos") of the set. The sequences of the capture oligonucleotides, immobilized on the support, are complementary and reverse (to allow hybridization with the RC oligos) and include an oligo T tail linker for immobilization.
a - numbers at the end of the probe names refer to their relative positions on the E. coli gyrB gene b - M. avium, M. intracellular, M. maloense c - M. tuberculosis, M. bovis, M. gastri
Oligonucleotides for immobilisation were custom synthesised (VBC Genomics, Vienna, Austria) with a 5' aminolink followed by five thymidine residues preceding the probe sequence. A 384 well flat bottom plate (Ritter GmbH, Schwabmunchen, Germany) was prepared with 30 μl of 50 μM oligonucleotide solution in Arraylt spotting buffer (TeleChem Inc., Sunnyvale, CA, USA). Microarrays were spotted with an OmniGrid spotter (1 TeleChem SMP3 pin) at • 55% relative humidity and 21°C onto aldehyde silylated slides (CEL Associates Inc., Pearland, TX, USA) . Arrays were spotted in triplicate to allow a statistical correction of the errors. Slide processing was carried out as described before (Stralis- Pavese et al., 2004). Processed slides were stored desiccated at room temperature in dark.
DNA templates
Strains used for microarray validation are listed in Table 2. Genomic DNA samples from pure cultures were purified using the DNeasy extraction kit (QIAGEN, Hilden, Germany) , following manufacturer's instructions. Archived veterinary samples used for the testing of the method' s performance were kindly provided by Clyde Hutchinson (Institute of Zoology, London, UK) . These DNA samples were prepared from animal specimens known to be contaminated with one or more bacterial pathogens using the DNeasy extraction kit (QIAGEN, Hilden, Germany) . Genomic DNA from soil samples was extracted using UltraClean Soil DNA Kit (MO BIO Laboratories, Carlsbad, CA, USA) following the manufacturer's instructions .
Table 2: List of bacterial species used for microarray validation. Strains were environmental or clinical isolates from the culture collection of the University of Sassari, Department of Biological Sciences, except for Pseudomonas putida, which was obtained from DSMZ (German Collection of Microorganisms and Cell Cultures) .
gyrB ac¬
Name Designation cession numbers
Aeromonas hydrophila SSM 4131 DQ386876
Campylobacter jejuni SSM 4129 -
Enterobacter cloacae SSM 2554 DQ386885
Escherichia coli SSM 1771 DQ386875
ETEC (E. coli enterotoxic) SSM 4135 DQ386871
EHEC slt-I (E. coli enterohaemorragic) SSM 4134 DQ386872
EHEC slt-II -(E. coli enterohaemorragic) SSM 4140 DQ386873
EIEC (E. coli enteroinvasive) SSM 4137 DQ386888
EPEC (E. coli enteropathogenic) SSM 4136 DQ386874
Helicobacter pylori SSM 4138 DQ386880
Mycobacterium avium SSM 4139 DQ386879
Mycobacterium tuberculosis
Proteus mirabilis SSM 2105 DQ386881
Pseudomonas spp SSM 2506 DQ386884
Salmonella spp SSM 1592 DQ386877
Serratia marcescens SSM 2560 DQ386886
Shigella spp SSM 2023 DQ386882
Staphylococcus aureus SSM CI-I DQ386887
Vibrio cholerae SSM 2508 DQ386878
Yersinia enterocolitica SSM 4130 DQ386883
DNA amplification gyrB gene was amplified using universal primers UPl ( 5 ? -GAAGTCATCATGACCGTTCTGCAYGCNGGNGGNAARTTYGA-S ' ) and UP2r ( 3 ' -TACTGNCTRCGNCTRCANCTRCCGAGCGTGTAGGCATGGGACGA-S ' ) . For several strains alternative primers UPlG (5'-GAAGTCATCATGACCGTTCTG- CAYGCNGGNGGNAARTTYGG-3' ) and UP2Ar
(3 ' -TACYGNCTRCGNCTRCANCTRCCGAGCGTGTAGGCATGGGACGA-S ') had to be used. PCR reactions were performed in 100 μl aliquots, consisting of Ix PCR buffer, 2 mM MgCl 2 , 4U Taq DNA polymerase (Invitro- gen, Carlsbad, CA, USA), 50 μM for each of the four dNTPs, 150 nM for each primer, with 50 - 100 ng DNA as template. For soil DNA and DNA mixtures amplification was performed with all four primers in one "multiplex" reaction, using the FailSafe PCR Pre- Mix E (Epicentre, Madison, WI, USA) . Amplification conditions were 95°C for 5 min, followed by 35 cycles of: 1 min at 95°C, 1 min at 58 0 C, 2 min at 72°C, followed by a final elongation step
of 10 min at 72 0 C
(http://seasquirt.mbio.co.jp/icb/protocols/protocols.php) . PCR products were subsequently purified using a commercial PCR purification kit (QIAGEN, Hilden, Germany) according to manufacturer's instructions and eluted in 30 μl dH 2 O. DNA concentration was measured using a NanoDrop spectrophotometer (NanoDrop Technologies, Wilmington, DE, USA) and adjusted to a final concentration of 50 ng/μl.
Cloning and sequencing
Purified PCR products were cloned using the TOPO TA Cloning Kit for Sequencing (Invitrogen, Carlsbad, CA, USA) following manufacturer's instructions. 25 μl of the chemically competent cells (provided with the cloning kit) were transformed with 6 μl of the ligation reaction. 100 μl of the transformation reaction were plated onto LB plates containing 100 μg/ml ampicillin and incubated overnight at 37 0 C. Template for the colony PCR was prepared by resuspending single colonies in 20 μl dH 2 O and denaturing the cell suspension for 10 min at 95°C in a thermocycler with subsequent cooling down to 4 0 C. 1 μl of this suspension was used as a template for 50 μl PCR reactions. M13 PCR reactions were performed as described above, but using M13 forward (5' GTAAAACGACGGCCAG 3') and M13 reverse primer (3' GTCCTTTGTCGATACTG 5') and 50 0 C annealing temperature. PCR products were subsequently purified using PCR purification kit (QIAGEN, Hilden, Germany) according to manufacturer's instructions and eluted in 30 μl dH 2 O. DNA concentration was measured using a NanoDrop spectrophotometer (NanoDrop Technologies, Wilmington, DE, USA) . Sequencing was performed using the BigDye Terminator vl.l Cycle Sequencing Kit (Applied Biosystems, Foster City, CA, USA) . Sequencing reaction was done in 10 μl reaction aliquots consisting of 4 μl purified PCR product, 2 μl sequencing mix and 400 nM primer (M13 forward or M13 reverse) . Sequencing conditions were 95 0 C for 2 min, followed by 25 cycles of: 30 sec at 95°C, 15 sec at 5O 0 C, 4 min at 60 0 C with no final elongation step. Sequencing reactions were purified using Sepha- dex G-50 (Amersham Biosciences, Piscataway, NJ, USA) columns and run on the automated ABI PRISM 3100 Genetic Analyzer (Applied Biosystems, Foster City, CA, USA) . Sequences were analysed using Sequencher v 4.5 (Gene Codes Corporation, Ann Arbor, MI, USA)
- -
and subjected to a preliminary nucleotide - nucleotide BLAST analysis within the NCBI database
(http://www.ncbi.nlm.nih.gov/BLAST) prior to importing to the gyrB database in ARB where proper phylogenetic analysis took place. Accession numbers of the gyrB sequences of the strains used in this study are listed in Table A.
Control pmoA PCR product
An internal positive control, a pmoA PCR product from Methylosinus trichosporium 0B3b, was included in each labelling and hybridisation experiment and subsequently applied for normalisation of the results. Genomic DNA extraction and pmoA PCR amplification were performed as described before (Bodrossy et al, 2003) . PCR product was purified using PCR purification kit (QIAGEN, Hilden, Germany) according to manufacturer' s instructions . DNA concentration was measured using a NanoDrop spectrophotometer (NanoDrop Technologies, Wilmington, DE, USA) and adjusted to a final concentration of 50 ng/μl.
DNA labelling
For the sequence-specific end labelling of the oligonucleotides a modified protocol of Rudi and co-workers (2003) was applied. After purification, gyrB PCR products were treated with shrimp alkaline phosphatase (SAP) (Roche Diagnostics GmbH, Pen- zbeg, Germany) in order to dephosphorylate remaining nucleotides. Samples consisting of 20 μl purified PCR product (50 ng/μl) , 2 μl Thermo Sequenase reaction buffer (Amersham Bios- ciences, Piscataway, NJ, USA) and 4 μl SAP (1 U/μl) were incubated at 37 0 C for 30 min, followed by 10 min at 95 0 C to inactivate enzyme activity. The SAP treated gyrB PCR products were then used for the cyclic labelling reaction. The control pmoA PCR product was further diluted to a final concentration of 5 ng/μl.
For the labelling a set of reverse complement oligonucleotides, lacking the 3' terminal cytosine residue, was custom synthesised (VBC Genomics, Vienna, Austria) . Lyophilised oligonucleotides were dissolved to a final concentration of 100 pmol/μl and stored at -20 0 C. For the labelling reaction an oligonucleotides mix pRC mix") , containing each reverse complement oligonucleotide at a final concentration of 1 pmol/μl, was pre-
pared. Competitive oligonucleotides, when used, were added at the same concentration to the RC mix.
The cyclic labelling was performed in 10 μl aliquots consisting of Ix Thermo Sequenase reaction buffer, 10 ng SAP treated control pmoA PCR product, 1 pmol of each reverse complement oligonucleotide, 10 pmol Tamra-ddCTP (PerkinElmer Life and Analytical Sciences, Boston, MA, USA) , 10 pmol of each ddATP, ddTTP, ddGTP (Roche Diagnostics GmbH, Penzbeg, Germany) , 3 U Taq DNA polymerase (Invitrogen, Carlsbad, CA, USA) or 3.2 ϋ Thermo Sequenase (Amersham Biosciences, Piscataway, NJ, USA) and 100 ng of SAP treated gyrB PCR product. Reaction conditions were 25 cycles of 30 s at 95°C followed by 75 s at 60 0 C, carried out in a thermocycler. After cyclic labelling, samples were used directly for hybridisation, without further purification. For experiments investigating the possibility of using up to 1000 different probes, RC oligos were added at a final concentration of 0.1 pmol/μl. Labelling experiments were also performed using all four nucleotides in form of Tamra-labelled ddNTPs (PerkinElmer Life and Analytical Sciences, Boston, MA, USA) at the final concentration of 10 pmol and omitting all silencing ddNTPs.
Enzyme inactivation
All experiments were performed by adding inactivating agent during cyclic labelling (10 μl reaction volume) and during hybridisation (210 μl reaction volume) . Reaction volumes were kept constant by adjusting dH 2 O volume added. Inactivation of the Thermo Sequenase was attempted with 2 U shrimp alkaline phosphatase, 100 mg Proteinase K and 1 M guanidine thiocyanate. Furthermore, purification of the labelled target was tested using phenol/chlorophorm method as well as Microcon YM-30 tubes (MiI- lipore Corporation, Billerica, MA, USA) .
Sensitivity tests
In order to establish the detection sensitivity of the mi- croarray analysis, experiments were performed with both artificially mixed samples (containing targeted strains in different ratios) and environmental samples spiked with 1% and 0.1% of the targeted strains.
Microarray hybridisation
Hybridisation was carried out as described before (Stralis- Pavese et al, 2004) . Labelled targets (10 μl) were mixed with 200 μl hybridisation buffer (pre-warmed to 65 0 C) . Final concentration of the hybridisation buffer was: β xSSC, ixDenhardt's reagent (Sigma, St. Louis, MO, USA), 0.1% SDS. Hybridisation was performed overnight at 55 0 C. After hybridisation, slides were washed in a 2xSSC, 0.1% SDS wash solution for 5 minutes, followed by two wash cycles for 5 minutes in 0.2xSSC, and a final wash for 5 minutes in 0. IxSSC, all at room temperature. Slides were dried with an oil-free air gun and scanned immediately.
Scanning and data analysis
Microarrays were scanned at three lines to average and at 10 μm resolution using a GenePix 4000A laser scanner (Axon Instruments, Foster City, CA, USA) . PMT gain was adjusted to scan the spots below the saturation level. Scanned images were saved as multilayer tiff images and analyzed with the GenePix Pro 6.0 software (Axon Instruments, Foster City, CA, USA) . Microsoft Excel was used for statistical analysis and presentation of the results. Microarray hybridisation results were normalised to the signal obtained from the internal control oligonucleotide (Msi_294) and expressed as percentage, 100% equalling the signal of the control probe. Probes were considered to be positive during validation if their normalised signal was at least 10% (of the control signal, Msi_294) .
Example 2: Results and discussion
Methodology adaptation
Sequence specific end-labelling of oligonucleotides (SSELO) (Rudi et al., 1998; Rudi et al., 2003; Rudi et al., 2002b), originally developed for membrane-based macroarrays, offers a significant improvement in the detection limit of MDMs by focusing labelling onto the regions actually used in hybridisation to the microarray. The principle of the method is shown in Figure 4. Capture oligonucleotides are immobilised on the microarray. Reverse complements of the capture oligos (RC oligos) are end labelled in a linear amplification reaction, upon the availability of the corresponding target sequence. The entire mixture is then
hybridised to the microarray to sort out the sequences which have been labelled. This way the amount of labelled nucleic acids, especially of those without a real capture probe on the micorarray, is drastically decreased. And thus is the level of nonspecific, background hybridisation, the major factor limiting detection sensitivity. Sequence specific end labelling of oligonucleotides (SSELO) was adapted to glass microarrays and for the application on relatively complex systems.
Using the Thermo Sequenase as in the original protocol, false positive signals were detected which apparently had nothing to do with the DNA sample used for labelling. By substituting the enzyme with its wild type ancestor, the Taq polymerase (Vander Horn et al., 1997) all these false positive signals disappeared without a compromise in signal intensity. Interestingly, all of the enzyme inactivation procedures tried failed at eliminating these false positive signals. Eventually, through a series of exclusion experiments it was determined that only four factors (i.e. Thermo Sequenase, Tamra-ddCTP, RC oligos and capture oligos) were essential for the occurrence of the false positive signals - not even the thermal cycling used in the labelling protocol or any template DNA was needed. If any of these four factors was absent, the hybridisation signal vanished. Without trying to find a thorough explanation for the phenomenon it has to be pointed out, that the Thermo Sequenase protocol was developed and successfully used on nylon membranes and amino microarray surfaces (Rudi et al., 2002a; Rudi et al., 2003) . The present experimental setup employs aldehyde microarray surfaces and the differences in the chemical and physical properties of the surface, the spatial organisation of the immobilised oligos and subsequent steric factors may be responsible for this observed effect.
The specificity of the assay was shown to be determined primarily by the stringency of the annealing step during labelling, rather than that of the subsequent hybridisation. The optimum annealing temperature under the conditions applied was shown to be 6O 0 C, below and above which false positive and negative results started to appear in increasing numbers. On the other hand, using targets labelled under optimised conditions, increasing the hybridisation temperature from the standard (55 0 C) to 60 0 C did not improve specificity further. In specific,
- -
the few false positive signals were not affected either. The only effect was an overall decrease in the intensity of all the signals .
One of the mechanisms conferring specificity to the labelling step is the application of labelled ddCTP in combination with the other three ddNTPs unlabelled (silencing ddNTPs) . This way the chance of an unexpected hybridisation of the RC oligo to a PCR product, yielding a labelled oligo is decreased by 75% (assuming a random chance of the next nucleotide in the 3' direction on the PCR product being G, serving as template for the addition of the labelled ddCTP) . However, this system seriously limits the options for probe design. This way one is allowed to select only probes where the nucleotide upstream of the RC oligo is a cytosine. Even though the present probe set was designed considering rule, it was checked whether relaxing it would compromise specificity. An environmental sample was labelled under standard conditions as well as in a modified reaction where all four dTNPs were labelled. Results performed using all four Tamra-ddNTPs and no silencing ddNTPs showed false positive results in only one out of 35 probes (i.e. Sal_1950) .
A crucial factor determining the potential performance of the method is the maximal number of probes that could be employed in parallel. The standard protocol applies 1 μl of a mixture of RC oligonucleotides with 1 pmol/μl final concentration for each oligo. Oligonucleotide stock solutions are usually 100 pmol/μl of concentration, further increase of which is quickly leading to an unstable solution, resulting in precipitation, uneven mixing, etc. Thus, at the amount of RC olgios applied, the maximal number of probes is limited in the range of 100-200 (our current array and hence the labelling procedure employs 35 oligonucleotide probes in one assay) . Decreasing the amount of RC oligos from 1 to 0.1 pmol did not, however, cause any significant drop in the signal intensities of the corresponding spots, indicating, that the method could be upgraded to accept up to 1000-2000 different oligonucleotides, subject to limitation only by the increasing number of potential interactions between the oligos .
Probe set validation
The probe set was validated with pure cultures of reference
strains (Figure 1) . Of the 35 probes validated only 1 (i.e. Prt_1882) exhibited false negative hybridisation results. This is most likely explained by the low Tm (48.3°C) being incompatible with the labelling temperature of 60 0 C. Several probes exhibited false positive signals with non-targeted strains exhibiting significant sequence similarity to the respective probes (calculated weighted mismatches from between 1.0 and 2.7). However, a competitive oligonucleotide system was developed that successfully resolved this problem. An overview of the validation data is presented in Figure 1.
Competitive oligonucleotides
In order to suppress false positive signals a series of competitive oligonucleotides ("CO oligos") was designed and tested. Competitive oligonucleotides were designed as a variation of an RC oligo showing false positive signals with non-targeted strains. The probe sequence was altered in a way to design a new probe that was a perfect match towards the strain exhibiting false positive signals. These CO oligos should therefore have a higher specificity towards the sequence giving rise to the false positive signal than the corresponding RC oligos. To ensure the silencing feature, CO oligos were synthesised with a 3' phosphate modification (VBC Genomics, Vienna, Austria) . CO oligos were included in the labelling reaction at the same concentration as the RC oligos. Of the 7 competitive oligonucleotides developed (Table 3) , 6 silenced false positive signal they were targeted against. Figure 3 shows hybridisation profiles of E. cloacae obtained with and without CO oligos. One of the competitive oligos (C0_l) failed to completely silence the false positive signal (i.e., below detection threshold). Still, a significant decrease in the signal was achieved. It has to be noted, that this was the strongest false positive signal obtained. Furthermore, this was the one resulting from the highest similarity between the RC oligo and the sequence giving rise to the false positive signal (weigthed mismatch value 0,7) - therefore having the lowest difference in the binding specificities for the CO and RC oligos. The application of the CO oligos had no statistically significant effect on the specific (i.e., expected) signals obtained with the targeted strains. Taken together, our results indicated that the application of
competitive oligonucleotides during the labelling step can significantly increase specificity without compromising sensitivity.
Table 3: Set of competitive oligonucleotides. Positions of mismatches with the original probe (RC oligo, see sequence in Table 1) are indicated by capital letters.
Probe set development
In this approach it is not the hybridisation on the array primarily determining probe specificity as every spotted probe encounters its perfect reverse complement match (either fluores- cently labelled or not) during it. This was also confirmed by obtaining virtually the same hybridisation profiles at 55 0 C and 60 0 C hybridisation temperatures (as detailed above) . The specificity conferring step of the assay is the cyclic labelling, the extension of the RC oligos by a single fluorescently labelled dideoxy nucleotide. Therefore the most crucial region is the 3' end of the oligo, followed by the middle. This is reflected in the parameters for calculating weighted mismatches, significantly differing from the parameters used for the approach where the hybridisation on the array is the step determining specificity (Stralis-Pavese et al., 2004).
Therefore, essential requirements during probe design are to tune the probe set in a way that their behaviour is compatible with the conditions of the cyclic labelling reaction and that probes have sufficient discriminating power. Three essential guidelines for the probe design were defined: i) position of the diagnostic mismatch (es) as close to the 3' end as possible, ii) similar melting temperature (60 ± 2 0 C, if possible) and iii) similar probe length (17-28 nt) . Probes potentially forming hairpin structures or 3' self-dimers that could lend themselves
to self-priming were avoided. Probe-target pairs with weighted mismatch (wMM) values of up to 0.5 were expected to yield positive hybridisation under the conditions applied. Seeing as most of the developed probes exhibited expected hybridisation results, the original guidelines established for probe design were indeed appropriate for this application. In fact the only probe exhibiting false negative results was the one most divergent from these requirements (Prt_1882, Tm=48.3°C).
However, in silico predictions for specificity could not always be confirmed experimentally. Several probes with wMM values between 0.7 and 2.7 exhibited unexpected ("false") positive signals. Parameters for mismatch weighing can without doubt be refined by considering these results and further ones. This will lead to a higher success rate of in silico specificity prediction.
Sensitivity
The sensitivity of MDMs is usually defined as the lowest relative abundance of the target group detectable (within the analysed community) . When reported, it is found to be around 5% (Loy et al., 2002; Bodrossy et al., 2003; Tiquia et al., 2004; Denef et al., 2003) . For short oligoprobe arrays this limitation is due to the target consisting of labelled nucleic acid fragments (several hundred nucleotides long) , spanning the whole target gene amplified, increasing the potential for the accumulation of background signal arising from a low rate of non-specific hybridisation. In practice, using these "conventional" approaches, each probe develops a low background signal of around 1-5% in respect of the maximal signal obtainable (Bodrossy & Sessitsch, 2004) . By focusing labelling only onto the region targeted by the actual probes, the SSELO approach minimises this background signal.
Experiments with environmental DNA from regular garden soil, apparently free of the bacteria targeted by our microarray, showed hardly any detectable signal on the probes of the array. Careful analysis showed that the highest signal detected on any of the probes was still less, than 0.1% of that obtained during validation (i.e., maximal signal of the probe) . Spiking the same soil DNA with 0.1% of Vibrio cholerae DNA resulted in clearly detectable signals for 2 of the 4 V. cholerae probes (Figure 2) .
_ -
Similar results were obtained with E. coli mixed into soil DNA as well as with artificial mixtures of pure genomic DNAs.
During array validation with pure cultures, some instances of detectable cross-hybridisations were detected, most of which were addressed via the application of competitive oligos . However, this raises the possibility of mistaking such weak cross-hybridisations with signals from low abundance bacteria. Such signals arose, however, only during validation and when environmental samples were spiked at a reasonable (i.e., >1%) ratio with a pathogen - i.e., when a targeted sequence was present at a high abundance. The weak false positive signals in these cases were due to the probe (s) specific to that high abundance pathogen, labelled in a specific manner, but displaying slightly non-specific hybridisation. Such potential uncertainties can be solved by repeating the labelling and hybridisation, but dropping the RC oligos specific for the pathogen (s) deteceted in high abundance, from the labelling mixture. Therefore, for environmental applications, a two step approach is suggested, first analysing whole community DNA with a complete set of probes, and subsequently re-analysing the sample with a probe set excluding probes for highly abundant species .
The sensitivity of the analysis is also influenced by the inherent bias in the PCR amplification of the gyrB gene.
Proof of concept
The applicability of the microarray was also demonstrated by analysing two archived veterinary samples (courtesy of Clyde Hutchinson, Institute of Zoology, London, UK) . The two samples were pathological stool samples from a harbor porpoise and a greenfinch, both having shown pathological signs of Salmonella infection. The microarray indicated the presence of Salmonella in both samples, as well as a lower, but clear signal for Vibrio cholerae. Unfortunately the second finding could not be confirmed by cultivation based experiments as the storage of the samples did not allow for the long-term survival of bacteria. However, Vibrio signals were never found in any of the hybridisations with reference strains and control hybridisations with environmental DNA strongly supporting the case that indeed a mixed infection with Salmonella and Vibrio was present in both samples. Even these preliminary results indicate the potential
of parallel screening for multiple pathogens in delivering new, precious information on various aspects of the ecology, diversity, interactions and epidemiology of bacterial pathogens.
Conclusions
Methods for the detection and identification of pathogenic bacteria need in most cases be very specific and highly sensitive. Combination of a unique labelling method (SSELO, sequence specific end labelling of oligonucleotides) , a housekeeping gene with a robust phylogenetic resolution at the species level (gyrB) , microarray technology and the concept of competitive oligos during labelling resulted in a highly specific method with significantly enhanced sensitivity (0.1% of the total microbial community analysed) , capable of addressing very complex microbial communities.
References
Bae et al. 2005. Nucleic Acids Res. 33: ell3.
Bodrossy et al. 2004. Current Opinion in Microbiology 7: 245-
254.
Bodrossy et al. 2003. Environmental Microbiology 5: 566-582.
Denef et al . 2003. Environmental Microbiology 5: 933-943.
DeSantis et al. 2005. FEMS Microbiol Lett 245: 271-278.
Hashsham et al. 2004. Biosens.Bioelectron. 20: 668-683.
Lee et al. 2005. J.Microbiol. Methods 61: 87-94.
Lehner et al. 2005. FEMS Microbiology Letters 246: 133-142.
Leinberger et al. 2005. Journal of Clinical Microbiology 43:
4943-4953.
Loy et al. 2006. Clinica Chimica Acta 363: 106-119.
Loy et al. 2002. Applied and Environmental Microbiology 68:
5064-5081.
Ludwig et al. 2004. Nucleic Acids Res. 32: 1363-1371.
Maynard et al. 2005. Applied and Environmental Microbiology 71:
8548-8557.
Neufeld et al . 2006. Environmental Microbiology 8 : 126-140 .
Rudi et al . 2002a . Applied and Environmental Microbiology 68 :
1146-1156 .
Rudi et al. 1998. Applied and Environmental Microbiology 64:
2639-2643.
Rudi et al. 2002b. Bio/Techniques 33: 496, 498, 500.
Rudi et al. 2003. ScientificWorldJournal. 3: 578-584.
SantaLucia et al. 1996. Biochemistry 35: 3555-3562.
Schadt et al. 2005. Methods In Microbiology 34: 332-368.
Sergeev et al . 2004. Biosens.Bioelectron. 20: 684-698.
Stralis-Pavese et al. 2004. Environmental Microbiology 6: 347-
363.
Tiquia et al. 2004. Bio/Techniques 36: 664-665.
Vander et al. 1997. Bio/Techniques 22: 758-5.
Wang et al. 2005. Journal of Clinical Microbiology 43: 4121-
4128.
Yamamoto et al. 1995. Appl. Environ Microbiol 61: 1104-1109.
Zhou 2003. Current Opinion in Microbiology 6: 288-294.